The Pochettino press

One of the best insights into Mauricio Pochettino’s tactical philosophy is given by the Southampton midfielder Morgan Schneiderlin in an interview with French newspaper L’Équipe. Asked about the pressing style at Southampton under Pochettino, Schneiderlin stated:

“First of all, we have to close the central zones. He keeps saying it’s the heart of play and there’s always more options from the central zones: switches of play etc… He emphasises blocking passing lanes. Anyway when I do press, I try to leave the worst passing option possible to the opponent. Pochettino asks us not to give the opponent the choice… He wants us to recover the ball as high as possible, so for that it’s usually up to a forward to trigger the pressing.”

The emphasis on putting pressure on the opposition team in their own half in order to disrupt their attacking moves and to recover the ball, is frequently reiterated by both Pochettino and those who have played under him, and there can be little doubt that this is how, in theory at least, he wants his teams to play. In a discussion about his style at Southampton, for example, he noted:

“Our style of play is to win back the ball as soon as possible and then play it. We moved forward our lines and play more upfield. When we lose the ball we must have the mentality of winning it back as soon as possible. It may seem like we are running more but really we are just running in a more organised way.”

There are several metrics which can help to measure the effectiveness of a pressing style of play, and one writer in particular has both attempted to create a statistical model to judge the effectiveness of a team’s press, and has written an article on the differences between Pochettino’s and AVB’s version of the high press.

In this article, however, I wish to focus more narrowly on just one metric for judging the high press: namely, the frequency of successful tackles within the opposition half (henceforth, OTs). There are a few reasons for this choice. First, it is a relatively easy statistic to measure, and, even without access to Opta, I have been able to look at match action zones on Squawka to see where tackles took place for the games under discussion.

Second, while a couple of articles have already been written about Tottenham’s greater propensity to make tackles in the opposition half this season, as far as I can tell no writer has looked at the effectiveness of individual Tottenham players within this system.

Finally, as Michael Caley has recently made clear in an extremely interesting analysis for The Washington Post, tackles in the opposition half are one of the most effective ways to create attacking chances. Tackling in the opposition half is – perhaps counter-intuitively – one of the best statistics for measuring how successful Tottenham’s attacking play has been this season.

I shall begin with a graph showing Tottenham’s OTs made during the 35 Premier League matches played so far this season. As can be seen, there are a few peaks and troughs as the Spurs team adapted to the pressing system, culminating with a nadir of 0 OTs in the away game at Aston Villa. Shortly after that, however, the frequency of successful tackles increases and stays at a largely consistent level until the home game against Swansea. It seems likely that the heavy work-rate required for a pressing game, coupled with Pochettino’s reluctance to heavily rotate the first XI, caught up with the squad by that point, and the rate of OTs has fallen away thereafter.

The first question that must be asked of course, is whether this OT rate is actually any better than in previous seasons?

In order to answer this question, I have plotted a graph showing the successful opposition half tackles made by Tottenham in the three previous seasons.

Impressionistically, the graph seems to suggest a gradual increase in OTs over the last three seasons, and the overall totals back this up. So far this season Tottenham have made 164 OTs at a rate of 4.7 per game. This is up on the 154 OTs (from 38 games) at a rate of 4.1 per game in the 2013-14 season, which itself is an increase on the 136 OTs at a rate of 3.6 per game achieved in 2012-13.

Indeed, not only has Pochettino got Tottenham performing OTs more often than in previous seasons, he has also got Spurs outperforming his Southampton team of 2013-14. That team managed 4.1 OTs per game.

Given Pochettino’s largely static tactical philosophy, the best way of explaining the difference between the frequency of OTs performed by both teams is to suggest that the Spurs players have adapted better to the pressing game than their Southampton counterparts. This is true to an extent, but an analysis of both the total OTs made by a number of Tottenham players, and their minutes per OT ratio, reveals some extremely interesting information. For the sake of convenience, I have analysed the players according to their position (centre backs have been omitted because of the infrequency with which they perform OTs), and have provided a brief overview for each category.

One of the most interesting things suggested by the statistics for the Tottenham forwards, is that the overall team may actually have adapted fairly quickly to the pressing game – it is simply that the wrong personnel started at the beginning of the season. In fact, there is a direct correlation between Harry Kane’s introduction into the first XI and an improved rate of team OTs.

If you look at the graph for the Tottenham 2014-15 season above, you will see that the OT rate only becomes consistently high from the time of the match at home to Stoke. This match also marks Kane’s first full 90 minutes in the Premier League. This correlation should underline Kane’s key role not just as a goal-scoring threat, but as a fundamental component of the Pochettino pressing system.

Interestingly, this is also probably the key explanation for how Pochettino has achieved a higher OT rate at Tottenham than at Southampton. Looking at the data for the 2013-14 season, Rickie Lambert achieved just 5 opposition tackles at a rate of 1 every 564 minutes, while Pablo Osvaldo did a little better (but still worse than Adebayor!), with 4 tackles at a rate of 1 every 214 minutes. Indeed, the problems the Southampton strikers had adapting to the pressing game was broached in the Schneiderlin interview. When asked about it he replied:

“We want to play with a high pressure, and those two are not famous for their stamina… They do not like to run after the ball, that’s not the way they play at all. They were both not putting in the effort to get the ball back, and it did not fit with our system”.

While Lennon and Townsend have excellent tackling statistics, it is probably worth bearing in mind that it is arguably easier for substitutes to make tackles against a tiring opposition. As such I’m not entirely convinced that their numbers would be replicated over a longer stretch of game time.

The number which really stands out here is that of Nacer Chadli. There has been a great deal of general grumbling about Chadli’s contribution to team play despite his goal-scoring prowess this season, and this figure would seem to bear it out. In fact, the fact that Jay Rodriguez managed not only 15 goals and 3 assists, but also 14 OTs at a rate of one every 184 minutes, probably explains why Pochettino seems to be so fixated on him. The fact that he can – or at least could, before his ACL injury – offer the same goal-threat as Chadli, while actually being able to play within the pressing system suggests that he would be a much better fit for the Spurs first XI.

While I suggested above that the OT rate for substitutes is likely to be deceptively high, it still seems bewildering that Stambouli has not been given more appearances in the Premier League. His incredible rate of an OT every 58 minutes is better than that of all Tottenham and Southampton players under Pochettino (including those who were mainly used as substitutes), and his hustling style of play also seems closely aligned with Pochettino’s philosophy. I would suggest that he has not been given a fair chance alongside Nabil Bentaleb in the Tottenham midfield, and it might be worth giving him some consistent game time to see if he can make an impression before the summer transfer window.

Mason deserves credit for his ability to make OTs, although it might be suggested that his slightly more advanced role – vis-à-vis Bentaleb – helps to account for his higher success rate. Both of Tottenham’s regular central midfielders fall short of their 2013-14 counterparts at Southampton, however, where Schneiderlin managed 22 tackles at an average of 1 every 126 minutes, and Jack Cork achieved 18 tackles at an incredible 1 every 96 minutes. It is somewhat surprising that Pochettino did not bid for Cork when he joined Swansea for £3.5m in January, and it seems likely that Tottenham will be looking to recruit at least one new central midfielder in the summer.

The stand out full back here is clearly Danny Rose, and his ability to regularly achieve OTs is one of the many reasons he has been first choice ahead of Ben Davies this season. The most worrying aspect of this set of statistics is how weak Kyle Walker has been in this regard. While it is true that his season has been heavily disrupted by injury, his numbers are very low and Tottenham must make sure that they have someone in the right back role capable of matching Rose’s success.

In terms of adapting the Tottenham team to meet the needs of the high press then, it is clear from the analysis of individual player statistics that Spurs require at least a back-up striker, a left attacking midfielder, a central midfielder and a right-sided full back in order to build a squad capable of properly implementing the Pochettino style.

A broader question, and one I can only begin to discuss here, is how the number of opposition half tackles actually correlates with the win rate. I am sure that someone with more statistical nous than me can come up with a more comprehensive answer, but I do have two contributions to make to this question.

First, I have also produced a graph showing the OT rate of Chelsea this season. Chelsea are not particularly known for playing a pressing game, yet their average of 4.3 OTs per game is not much lower than Tottenham’s. I would tentatively suggest that one of the reasons Chelsea’s rate is so high might be because of the tendency of their opponents to drop extremely deep. If the opposition play largely within their own half then the OT rate is clearly going to be higher. Moreover, in this scenario a successful OT might not be as dangerous because the opposition could still have the majority of their players behind the ball.

I leave it to people with access to better data than me to check (a) whether Chelsea’s opponents tend to play deeper than Tottenham’s, and (b) whether the number of opposition half tackles leading to goal scoring opportunities is higher for Tottenham or Chelsea. One thing that is clear from the Chelsea graph, however, is that there is more match-to-match variation in their OT rate than there is with Tottenham or Southampton.

Second, if the Tottenham average tackle rate this season is 4.7 per game, then there might be some value in seeing the number of games won, lost and drawn when Tottenham fall short of and exceed this number. The results are as follows:

As can be seen, there is not a great disparity between results when Tottenham’s OTs are below average and those when they are above average. Indeed, Tottenham won the only match of the season when their OT rate was 0 (away at Aston Villa), and have lost 2 games despite achieving 7 OTs (home to Stoke and away to Liverpool).

Of course, as I mentioned at the beginning of this article, tackles in the opposition half is only one metric for gauging the effectiveness of a high press, and it may be the case that a more holistic statistical overview will yield a better correlation. For now, however, I want to finish by suggesting that while Mauricio Pochettino has made a good initial start in improving Tottenham’s OT rate, he still has some way to go to create a team in the mould of the true pressing behemoths. Below is a graph comparing Tottenham with the Dortmund team of 2014-15. That Dortmund team had an OT rate of 5.6 per game, and this is clearly the kind of level Tottenham should be aspiring to reach.